Feature selection via maximizing neighborhood soft margin

Qinghua Hu; Xunjian Che; Jinfu Liu

doi:10.1007/978-3-642-05224-8_13

Feature selection via maximizing neighborhood soft margin

Qinghua Hu^*
, Xunjian Che
, Jinfu Liu

^*Corresponding author for this work

School of Energy Science and Engineering

Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Feature selection is considered to be a key preprocessing step in machine learning and pattern recognition. Feature evaluation is one of the key issues for constructing a feature selection algorithm. In this work, we propose a new concept of neighborhood margin and neighborhood soft margin to measure the minimal distance between different classes. We use the criterion of neighborhood soft margin to evaluate the quality of candidate features and construct a forward greedy algorithm for feature selection. We conduct this technique on eight classification learning tasks. Compared with the raw data and other three feature selection algorithms, the proposed technique is effective in most of the cases.

Original language	English
Title of host publication	Advances in Machine Learning - First Asian Conference on Machine Learning, ACML 2009, Proceedings
Pages	150-161
Number of pages	12
DOIs	https://doi.org/10.1007/978-3-642-05224-8_13
State	Published - 2009
Event	1st Asian Conference on Machine Learning, ACML 2009 - Nanjing, China Duration: 2 Nov 2009 → 4 Nov 2009

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	5828 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	1st Asian Conference on Machine Learning, ACML 2009
Country/Territory	China
City	Nanjing
Period	2/11/09 → 4/11/09

Access to Document

10.1007/978-3-642-05224-8_13

Cite this

Hu, Q., Che, X., & Liu, J. (2009). Feature selection via maximizing neighborhood soft margin. In Advances in Machine Learning - First Asian Conference on Machine Learning, ACML 2009, Proceedings (pp. 150-161). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5828 LNAI). https://doi.org/10.1007/978-3-642-05224-8_13

@inproceedings{a963eb917a8b4e83af7dc1736867d8ce,

title = "Feature selection via maximizing neighborhood soft margin",

abstract = "Feature selection is considered to be a key preprocessing step in machine learning and pattern recognition. Feature evaluation is one of the key issues for constructing a feature selection algorithm. In this work, we propose a new concept of neighborhood margin and neighborhood soft margin to measure the minimal distance between different classes. We use the criterion of neighborhood soft margin to evaluate the quality of candidate features and construct a forward greedy algorithm for feature selection. We conduct this technique on eight classification learning tasks. Compared with the raw data and other three feature selection algorithms, the proposed technique is effective in most of the cases.",

author = "Qinghua Hu and Xunjian Che and Jinfu Liu",

year = "2009",

doi = "10.1007/978-3-642-05224-8\_13",

language = "英语",

isbn = "3642052231",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "150--161",

booktitle = "Advances in Machine Learning - First Asian Conference on Machine Learning, ACML 2009, Proceedings",

note = "1st Asian Conference on Machine Learning, ACML 2009 ; Conference date: 02-11-2009 Through 04-11-2009",

}

Hu, Q, Che, X & Liu, J 2009, Feature selection via maximizing neighborhood soft margin. in Advances in Machine Learning - First Asian Conference on Machine Learning, ACML 2009, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5828 LNAI, pp. 150-161, 1st Asian Conference on Machine Learning, ACML 2009, Nanjing, China, 2/11/09. https://doi.org/10.1007/978-3-642-05224-8_13

Feature selection via maximizing neighborhood soft margin. / Hu, Qinghua; Che, Xunjian; Liu, Jinfu.
Advances in Machine Learning - First Asian Conference on Machine Learning, ACML 2009, Proceedings. 2009. p. 150-161 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5828 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Feature selection via maximizing neighborhood soft margin

AU - Hu, Qinghua

AU - Che, Xunjian

AU - Liu, Jinfu

PY - 2009

Y1 - 2009

N2 - Feature selection is considered to be a key preprocessing step in machine learning and pattern recognition. Feature evaluation is one of the key issues for constructing a feature selection algorithm. In this work, we propose a new concept of neighborhood margin and neighborhood soft margin to measure the minimal distance between different classes. We use the criterion of neighborhood soft margin to evaluate the quality of candidate features and construct a forward greedy algorithm for feature selection. We conduct this technique on eight classification learning tasks. Compared with the raw data and other three feature selection algorithms, the proposed technique is effective in most of the cases.

AB - Feature selection is considered to be a key preprocessing step in machine learning and pattern recognition. Feature evaluation is one of the key issues for constructing a feature selection algorithm. In this work, we propose a new concept of neighborhood margin and neighborhood soft margin to measure the minimal distance between different classes. We use the criterion of neighborhood soft margin to evaluate the quality of candidate features and construct a forward greedy algorithm for feature selection. We conduct this technique on eight classification learning tasks. Compared with the raw data and other three feature selection algorithms, the proposed technique is effective in most of the cases.

UR - https://www.scopus.com/pages/publications/70549105003

U2 - 10.1007/978-3-642-05224-8_13

DO - 10.1007/978-3-642-05224-8_13

M3 - 会议稿件

AN - SCOPUS:70549105003

SN - 3642052231

SN - 9783642052231

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 150

EP - 161

BT - Advances in Machine Learning - First Asian Conference on Machine Learning, ACML 2009, Proceedings

T2 - 1st Asian Conference on Machine Learning, ACML 2009

Y2 - 2 November 2009 through 4 November 2009

ER -

Feature selection via maximizing neighborhood soft margin

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this